[hts-users:04488] Re: Visualize data alignment

[Franco Negri <franconegri2004@xxxxxxxxx>]

Thanks you all for the help.

I'm using the EHMM automatic alignment present in festvox.

On Wed, Feb 15, 2017 at 2:19 PM, Luiz Felipe Santos Vecchietti
<lfelipesv@xxxxxxxxxxxx> wrote:
> Franco, which tool are you using to align your data? HTK?
>
> Cheers,
>
> Luiz Felipe.
>
> 2017-02-15 13:38 GMT-02:00 Rasmus Dall <R.Dall@xxxxxxxxxxxx>:
>>
>> Wavesurfer (http://www.speech.kth.se/wavesurfer/) can load HTK style .lab
>> files so if you have those you can use that.
>>
>> Also, for wavsurfer, you should convert the raw files to wav files first.
>>
>> - Rasmus
>>
>>
>> Quoting Erica Cooper <ecooper@xxxxxxxxxxxxxxx> on Wed, 15 Feb 2017
>> 10:30:09 -0500:
>>
>>> Praat (http://www.fon.hum.uva.nl/praat/) can be used for this purpose,
>>> only
>>> the label files must first be converted into the Praat TextGrid format.
>>> We
>>> have some scripts for doing that here (from the monophone .lab files):
>>> https://github.com/ecooper7/label-convert
>>>
>>> Best,
>>> Erica
>>>
>>> On Wed, Feb 15, 2017 at 10:00 AM, Franco Negri
>>> <franconegri2004@xxxxxxxxx>
>>> wrote:
>>>
>>>> Hello,
>>>> Recently I used the english demo to generate a new spanish voice.
>>>> Though the results are acceptable, they are not perfect.
>>>> I suspect this is because the .utt files for training are not
>>>> perfectly aligned with the .raw files. Do you know of any tool I can
>>>> use to visualize the alignment of a .utt (or .lab) file and it's
>>>> corresponding .raw?
>>>>
>>>> Regards,
>>>> Franco Negri.
>>>>
>>>>
>>>>
>>>
>>
>>
>>
>> --
>> The University of Edinburgh is a charitable body, registered in
>> Scotland, with registration number SC005336.
>>
>>
>>
>